Reliability and Accessibility of Low-Latency V2I Channel Training Protocol Using Cover-Free Coding: Win-Win or Tradeoff ?

Future vehicle-to-infrastructure communications in a multi-vehicle environment highly depend on the acquisition of wireless orthogonal frequency division multiplexed (OFDM) channels through the conventional pilot-tone based channel training protocol. However, this protocol can be easily compromised by a pilot-aware attacker spoofing/jamming/nulling those pilot tones, which have been designated for multiple vehicles. To solve this issue, we propose merging the multi-vehicle signal characteristic into the well-known cover-free coding technique for securely and retrievably coding pilot tones of multiple vehicles. In particular, multiple vehicles encode and convey their own pilot tones into diversified subcarrier activation patterns (SAPs). Then, road side unit observes the superposition of those encoded SAPs and re-designs the decoding procedure using signal independence characteristic such that those SAPs, though camouflaged by malicious signals in wireless environment, could be separated, identified, and decoded into original pilots. The protocol reliability is formulated using pilot identification error probability (IEP), and the protocol ξ-accessibility is defined by the ratio of the number of accessible vehicles to the consumed OFDM symbols under the constraint of IEP less than ξ. We prove that a win-win relationship between reliability and accessibility always exists and also demonstrate how more number of antennas of the attacker could improve the reliability better. Finally, numerical results verify how the proposed method is enabled to support ultra-reliable (1-ξ≥ 99.999%) and ultra-low-latency (up to 1.5 ms) multi-vehicle access for channel training under an Long Term Evolution (LTE) 20 MHz bandwidth.